Delphi Markets Inc.
A complete market is one where every future state of the world has a corresponding tradeable security.
The primary condition for such a market is the existence of Arrow securities. An Arrow security pays $1 if a particular state of the world occurs at a future date, and $0 otherwise. This condition is satisfied with the emergence of prediction markets as a new asset class.
The secondary condition is information symmetry: this ensures that uninformed actors are not at a permanent disadvantage. Information symmetry is not possible. The theory of fallibility states that “[market] participants’ view of the world is always partial and distorted”. Adverse selection (when one party knows more than the other) will always exist, in all asset classes. Though pricing will continue to improve as the new asset class calibrates, a certain level of information asymmetry is unavoidable. Markets are efficiently inefficient.
Fallibility guarantees a disconnect between price and reality. If participants’ views were in line with the true view of the world, price would always equal reality. Reflexivity is a property of markets that naturally emerges from this disconnect; the world state itself is partly shaped by the participants, even the uninformed.
Reflexivity is the relationship between prices and reality in a market. The existence of reflexivity in prediction markets is what makes a market a market, not a fancy forecasting platform. For example, if a public company’s stock price increases, they can now better allocate capital to improve their fundamental value (GameStop). Markets are engines, not cameras; they are endogenous to society. Pricing a risk may change the probability of that risk occurring, because perceptions and reality reinforce each other. Event contracts today are not reflexive — the price of a market is a forecast of an event determined outside the market (example: does seeing your preferred candidate’s odds change whether you vote for them?).
Forecasts conditioned on event forecasts are reflexive. For example, seeing that your preferred candidate being voted into office will significantly increase the cost of oil (and hence, gas) might impact your voting preferences.
Prediction markets are not complete markets. In fact, they are not markets. Reality should affect the market’s price, and the market’s price should affect reality. Prediction markets today satisfy the former. A complete market will price the aftermath. When consequences can be seen, decisions change. When decisions change, so does the future.
Welcome to Delphi.